Perlman - Practical Ecology for Planners, Developers and Citizens - Chapter 6 pdf

Imagine taking a flight across North America on a clear day—from, say, New
York to Vancouver—and describing the patterns you observe on the land below.
After lifting off, you would fly over industrial and residential landscapes criss-
crossed by numerous roads and broken up by the occasional park or greenway.
As you left the city behind, forests would begin to dominate, punctuated by farm
fields and towns. You might see patches of lighter and darker green, indicating
different forest types. Farmlands in the Midwest would appear as a rectilinear
grid delineated by roads and hedgerows, while fields in arid eastern Washington
watered by center-pivot irrigation might appear as series of green circles against
a tan background of scrubland. Approaching the West Coast, you might see a
checkerboard of clear-cuts within the old-growth conifer forest.
While these landscapes vary tremendously, all of them can be described as
aggregations of three basic elements: patches, corridors, and matrix. When the
landscape is viewed from the air, these become quite apparent, with corridors
linking discrete patches in a surrounding matrix (see Figures 6-1a through 6-1c).
This pattern of elements is one of the major organizing principles of landscape
ecology, a relatively new branch of ecology that helps us understand the form
and function of features on the landscape. Richard Forman, Michel Godron, and
others helped this field coalesce in the 1980s after earlier work by ecologists, ge-
ographers, and landscape planners in West Germany and the Netherlands in the
6
The Ecology of Landscapes
Figure 6-1a. This image shows a
large patch of forest plus a smaller
patch of developed land within a
matrix of agricultural land.
Figure 6-1b. In this photo, a forest
corridor stretches between two
patches of forest within a matrix of
unforested wetlands and farmlands.
Figure 6-1c. Here, small patches of

farmland are interspersed in a
forested matrix.
1960s and 1970s.
1
Forman’s 1995 book Land Mosaics provides a more recent syn-
thesis of the field of landscape ecology.
2
Landscape ecology examines how the spatial arrangement of land uses af-
fects their function for humans, other life forms, and abiotic processes. Since
planning and development are first and foremost about the arrangement of land
uses within a site or community, this is indispensable knowledge. Landscape
ecology also allows us to infer something about natural processes and biodiver-
sity protection issues even when we have little ecological data about the land-
scape or the species that reside there. Thus, its principles can allow planners and
designers to make useful generalizations or reasoned hypotheses in cases when
they must make decisions based on incomplete information (which is almost al-
ways). Finally, the concepts of landscape ecology can be used for almost any land-
scape (urban, forested, agricultural) and at almost any scale.
In addition to introducing terrestrial landscape ecology and its relevance to
the planning fields, this chapter surveys the other components of landscapes:
aquatic ecosystems and abiotic elements. We then integrate these concepts with
those in Chapters 4 and 5 to present the ideas of ecological integrity and sus-
tainability—big-picture perspectives that can guide planners and designers in
their local projects.
A Word about Scale
Planners and designers work at different scales and in different contexts. For ex-
ample, a planner may work at the state/provincial, county, municipal, or site level,
while a landscape architect might design a planting plan for a single lot or a de-
velopment plan for thousands of acres or hectares. Ecologists use a separate hi-
erarchy of scales based on biological, not political, organization. Even though

there is no “standard” size for biological elements such as habitats and commu-
nities, some generalizations are shown in Table 6-1.
Although the term landscape is often used colloquially with a variety of
meanings, landscape ecologists use it to refer to the area that one can see from a
mountaintop or an airplane—an area where a given combination of local ecosys-
tems or land uses is repeated in similar form, usually for tens of miles or kilo-
meters.
3
Examples of landscapes might include the suburbs around a major city,
an agricultural valley, or a tract of national forest that is managed differently
from surrounding lands. An ecoregion encompasses many different landscapes
that may be quite dissimilar from one another but that are united by common
environmental conditions (such as climate or surficial geology), species, and dis-
turbance processes.
4
Just as planners sometimes work across political boundaries,
such as when they work in a multitown watershed area, conservationists often
The Ecology of Landscapes 95
use landscapes and ecoregions—which typically cross political borders—as the
primary organizational boundaries for their work.
What is the most appropriate scale at which to plan? The answer, actually, is
“all of them.” As planners know, it is often possible to be most successful at a
small scale, where one wields the most authority and political power. However,
grand achievements usually result only from large-scale visions. This paradox,
of course, is the reason for environmentalists’ exhortation to “think globally,
act locally.” Effective conservation does not occur in a vacuum; instead, as em-
phasized in Chapter 1, each site (or development or habitat) should be considered
in relation to its context and at a variety of different scales. So, if you are a plan-
ner or designer, first select the scale at which you work from the “Political/
Jurisdictional” column of Table 6-1. Then move to the right and look up one row

and down one row. These are the ecological scales that should be considered, at a
minimum, during planning. In the words of landscape ecologist Richard Forman,
planning professionals should “think globally, plan regionally, and then act
locally.”
5
Conservation biologist Reed Noss explores the topic of scale in his article
“Context Matters: Considerations for Large-Scale Conservation,” arguing that
the selection of too narrow a context for biodiversity conservation may lead to
negative consequences.
6
He describes how the managers of the 563-acre (228 ha)
96 THE SCIENCE OF ECOLOGY
Table 6-1.
Scale and Context for Planning and Conservation
Scale Political/Jurisdictional Ecological
(size of landscape element) (planners, designers, (conservation biologists)
developers)
Less than 500 acres Lots, sites, districts, and zones Habitats
500 to 5,000 acres Sites, districts, and zones Ecosystems and communities
10s of square miles Cities and towns Ecosystems and communities
100s to 1,000s of square miles Counties and regions Landscapes
1,000s to 100,000s of States and provinces Ecoregions
square miles
9.5 million square miles North America Continent
(land area only)
57.4 million square miles Earth Earth
(land area only)
Sugarcreek Nature Reserve in Ohio increased habitat diversity and species rich-
ness within the reserve by replacing maturing forest, which is home to relatively
rare forest interior species, with more common habitat types. By reducing the

amount of rare maturing forest in the reserve, however, they hurt the cause of
biodiversity protection in the broader region.
Form and Function of Matrices, Patches, and Corridors
Imagine viewing your hometown as if you were a deer, an eagle, a tortoise, a sala-
mander, or a beetle. Where do you live? What do you eat? Do you need to travel
between different habitats, and, if so, how do you get from one to another? Who
is trying to eat you, and how do you avoid them? These questions will help us
examine how the arrangement of patches, corridors, and matrices on the land-
scape affect the species that inhabit them.
Animals have three different types of space needs: space for a home range,
migration, and dispersal. The home range is the area used by the animal for day-
to-day feeding and shelter. For some territorial animals, home range is exclusive,
such that only one individual (or pair, family unit, or allied group) of that species
occupies any habitat patch at any given time. But for most species, home ranges
can overlap. Most animals have a minimum home range requirement and can-
not survive long term if they lack this amount of suitable habitat. Migration is
seasonal movement from one habitat to another, usually along a latitudinal or
altitudinal gradient. Migrating animals require adequate habitat for each season
as well as a suitable conduit for migration. Finally, dispersal is movement beyond
the animal’s typical day-to-day or season-to-season movement patterns; it is re-
sponsible for establishing new populations of a species and for interbreeding be-
tween separate populations. While dispersal is not essential for the survival of
individuals, it is important for the long-term viability of populations and species.
Dispersal, like migration, requires that suitable conduits for movement be avail-
able. Dispersal is also important for plants and other stationary life forms.
Matrices
The matrix is the dominant land use type or ecosystem in any given land-
scape. Examples of matrices include corn and soybean fields in eastern Nebraska,
temperate rainforest in the Pacific Northwest, or housing subdivisions in sub-
urban Los Angeles. The matrix is usually the most extensive land use type (based

on area), but sometimes its dominance is the result of being the most intercon-
nected or most “influential” land use type. For example, in a suburbanizing re-
gion, urban development may constitute the matrix even though it covers only
The Ecology of Landscapes 97
40 percent of the landscape. This is because the urban areas are completely
interconnected by roads and exert strong influences on native ecosystems, which
have been relegated to residual patches. The matrix can change over time—for
example, from agriculture to urban at the edge of a sprawling metropolis, or from
old-growth forest to early successional forest in a landscape with extensive clear-
cutting. In these examples, what was formerly the matrix would become residual
patches or corridors (see Figure 6-2).
Patches
Patches are created by several different processes. The unaltered landscape is
naturally patchy because of environmental variability (different soils, microclimate,
and water availability) as well as disturbance processes, such as fire, flooding, and
windstorms. Humans create patches by developing small outposts in a natural
matrix, such as when a few farmsteads are cut in a large forested area, or by chang-
ing the matrix so that only remnants of natural habitat remain in a domesticated
landscape, such as bits of forest or prairie surrounded by cultivated fields.
98 THE SCIENCE OF ECOLOGY
Figure 6-2. In this part of the western United States, the matrix land cover used to be
scrub vegetation. In the lower part of the photo, the matrix is now an expanding urban
area (with a few small patches of scrub vegetation within the matrix), while in the upper
part of the photo, the matrix is still scrub with a few small patches of residential devel-
opment and forest.
patch size
The size of natural patches affects the number and abundance of species they
contain. Ecologists first noted this pattern in the early 1900s and developed
species-area curves to plot the relationship between patch size and number of
species (see Figure 6-3). In 1967, ecologists Robert H. MacArthur and Edward O.

Wilson provided a theoretical explanation for this pattern in their equilibrium
theory of island biogeography, which attempts to explain why certain oceanic is-
lands contain more species than others.
7
The theory proposes that the number of
species on an island represents an equilibrium between the number of new
species colonizing the island and the number of preexisting species going locally
extinct on the island. Islands situated near the mainland receive more immi-
grating species than do distant islands and thus tend to have more species. Simi-
larly, big islands can support larger populations of given species than small islands
can. These larger populations are less likely to go extinct over time, implying that
large islands can support more species.
During the 1970s, some biologists began to apply island biogeography theory
to the design of nature reserves, arguing that, all else being equal, large nature
reserves and reserves that are close to other reserves will contain more species
than small and isolated reserves. This is an intuitive idea, but a few caveats are
worth noting. First, patches of terrestrial habitat are not true islands. The sur-
rounding matrix matters greatly, because this matrix can either enhance species
immigration or accelerate extinction. Second, the number of species in a patch
depends not only on area but also on habitat type, habitat diversity (i.e., the num-
The Ecology of Landscapes 99
Figure 6-3. As shown on this graph, species diversity (on the vertical axis) increases
with patch size (on the horizontal axis), rapidly at first and then more slowly. Patch
size is not the only factor affecting species richness: some habitat types are inherently
more species rich than others, as the two different curves illustrate.
COVE HARDWOOD FOREST
SPRUCE-FIR ALPINE FOREST
Area of Habitat Patch
Number of Plant Species in Patch
ber of different niches available), disturbances, and other factors.

8
The general-
ized species-area curves shown in Figure 6-3 illustrate that species richness
can differ greatly by habitat type, even for two habitats occurring very near
each other.
Finally, the species-area curve is not always a smooth line but may contain
“threshold” points for different ecosystems. One important threshold in many
ecosystems is the minimum patch size that will support viable populations of
predators and large herbivores, which are often keystone species. A patch at least
this large may be necessary to preserve an essentially intact example of a par-
ticular ecosystem. Thus, while bigger is usually better, conservation planners
must also pay attention to habitat diversity, patch context, and size thresholds
for different ecosystems.
patch shape and edges
The term edge effect refers to the different processes that occur at the edge
of a patch versus its interior. For example, the portion of a forested tract adja-
cent to a suburban backyard would tend to be warmer and drier than the forest
interior because of sun and wind penetration from the open backyard. The yard
might contribute other influences as well, such as pesticides and fertilizers from
the lawn, introduced predators such as cats and dogs, noise, and invasive species
(see Figure 6-4). While there is no firm rule on how far edge effects extend, sev-
100 THE SCIENCE OF ECOLOGY
Figure 6-4. Different edge effects extend different distances from settled areas into
natural habitats. The length of the arrows indicates the relative distance that each ef-
fect extends. (Please note that this diagram is not to scale.)
eral studies offer insight. Microclimate effects—such as elevated wind speed, ele-
vated soil temperature, and reduced moisture—typically extend one-half to one
tree height (roughly 30 to 100 feet, or 10 to 30 meters) into a forested patch but
were found to extend as far as two to three tree heights (200 to 400 feet, or 60
to 120 meters) into conifer forests in the Pacific Northwest.

9
The extent of the
microclimate edge effect depends on the forest type, the amount of understory
vegetation, and the patch’s orientation relative to the wind and sun.
Patch edges also tend to have different species than patch interiors do. Edges
often have a high diversity of species but commonly favor adaptable generalist
species as well as multihabitat species that depend on resources on both sides of
the boundary. Examples of common North American edge species include white-
tailed deer, raccoon, and skunk, all of which can be found in suburban and agri-
cultural landscapes with abundant edge. By contrast, interior species are intol-
erant of edge conditions and human disturbances, or they require habitat
characteristics that are found only in interiors. Examples of forest interior bird
species in North America include the northern goshawk, ovenbird, and various
warblers and vireos.
10
The effect of edges on species distribution reveals an important tension
among differing habitat management goals. For hunters, edge habitat is often de-
sirable since many game birds and mammals are edge species. For this reason,
land managers seeking to improve hunting opportunities have sometimes pur-
posefully increased the amount of edge in a landscape by cutting or burning vege-
tation. However, edges tend not to contain rare or endangered species and also
tend to attract generalist predators, which have been blamed for reducing popu-
lations of many rare songbird species, among other animals.
11
The edge effect on
species distribution can extend for several hundred yards or meters from a for-
est edge.
12
The shape of patches allows us to infer much about their origin and function.
Some of these relationships have been studied and confirmed by ecologists, while

others are essentially working hypotheses. Rectilinear patches and edges are al-
most invariably created and maintained by humans, whereas natural edges tend
to be irregular, with curves and lobes. Initial studies suggest that curvilinear and
lobed boundaries tend to promote wildlife movement across boundaries (animals
often enter or exit a patch at one of the lobes), whereas straight boundaries pro-
mote movement along boundaries.
13
Round patches contain more interior habi-
tat and less edge habitat than do elongated or convoluted patches of the same
total area. However, lobed and elongated patches tend to be more heterogeneous
than compact ones, which may promote greater genetic diversity and better re-
sistance to pests and disease as a result of populations within the patch being par-
tially isolated from one another.
The Ecology of Landscapes 101
Considering all these factors, what patch shape and what types of edges are
optimal from a conservation standpoint? Maximizing native biodiversity requires
both edge habitat and interior habitat. However, since edge habitat is usually
abundant in human-influenced landscapes, the first priority for nature reserves
is generally to protect interior habitat. A round patch with few irregular edges
maximizes interior habitat area. Depending on the situation, this basic shape
might be optimized according to the factors discussed above. For example, if the
area is subject to disturbance processes, such as fire or pest outbreaks, the addi-
tion of lobes offers a “risk-spreading” benefit, reducing the chance that a distur-
bance event will affect the entire patch at once.
Corridors
Landscape ecologists use the term corridor generically to refer to any land
use that is long and relatively narrow and either connects two or more patches
or interrupts or dissects the matrix. Corridors run the gamut from fundamen-
tally natural habitat, such as a strip of forest along a river, to human creations,
such as roads, railroads, and pipelines.

Five major functions of corridors have been identified.
14
As habitats, most
narrow corridors of residual or planted vegetation (such as hedgerows or buffers
around a development site) are dominated by edge species that can tolerate inputs
and disturbances from the surrounding matrix. However, some corridors are intact
natural habitats, such as riparian or ridgeline ecosystems. Corridors act as a con-
duit for movement, not just for animals but also for plants, humans, water, sedi-
ment, and nutrients. To the extent that they help plants and animals move across
the landscape, corridors often can improve the viability of populations and con-
tribute to conservation efforts. While corridors may facilitate movement for some
species or materials, they may act as a filter or barrier to movement for others.
In this way, a corridor can reduce or eliminate interactions between individuals
on either side, creating separate populations or, in the case of people, distinct
neighborhoods. Finally, corridors can function as a sink or a source for animals,
plants, people, water, air, heat, dust, or chemicals. For example, windbreaks planted
in agricultural areas in the 1930s following the Dust Bowl function as a sink for
dust particles and often as a source for insect- and crop-eating animals.
Because corridors typically serve different combinations of functions for dif-
ferent species and processes, it is important to tailor the function of any proposed
corridor to the intended purpose. The most important factors influencing corri-
dor functions are width, connectivity, and heterogeneity. A corridor of natural
habitat that is tens or even a couple of hundred feet (tens of meters) wide will
be mostly edge and consequently will be used mostly by generalist species. To
allow movement by interior species and many large mammals, corridors must be
102 THE SCIENCE OF ECOLOGY
hundreds to thousands of feet (hundreds of meters) wide to provide adequate
buffering from the matrix and adequate long-term protection from distur-
bances.
15

The appropriate width of stream corridors is discussed on pages 200–1.
Connectivity must be evaluated not just spatially (i.e., whether the green ribbon
on the map is continuous) but also functionally for the purposes of moving a spe-
cific animal or substance.
16
Factors that have been demonstrated or are believed
to make corridors better for animal movement include few narrows or gaps, fairly
straight configuration, little environmental heterogeneity, little crisscrossing of
streams or roads, and shortness of length.
17
benefits of habitat corridors in fragmented landscapes
In the popular and semitechnical literature, such as the magazines and Web
sites of some conservation groups, corridors are sometimes presented as an answer
to most conservation problems. For example, the Web site of Ecotrust, a conser-
vation group based in Portland, Oregon, states that “wildlife corridors are nec-
essary because they maintain biodiversity, allow populations to interbreed, and
provide access to larger habitats.”
18
The typical argument for corridors goes like
this. Before human land uses, such as agriculture and urban areas, came to domi-
nate, the landscape consisted of large blocks of intact habitat that allowed or-
ganisms wide freedom of movement. Today’s patterns of human land use have
fragmented the landscape and cut off patches of native habitats from one another,
thus isolating small populations of organisms that were once part of larger popu-
lations. These small populations face an increased risk of extinction. The solution,
according to many conservation biologists, is to decrease isolation by retaining
(or creating) corridors that link patches of native habitat.
The value of corridors for biodiversity conservation is the subject of current
debate and research among ecologists. Thus far, scientific evidence for the effi-
cacy of corridors is limited, but at least a dozen studies offer observational and

experimental evidence that corridors facilitate movement and dispersal between
habitat patches.
19
Given the difficulty of conducting large-scale ecological experi-
ments, most of this evidence relates to plants and smaller animals (insects, birds,
and small mammals) on relatively small habitat patches. This, however, is the
scale at which most planners and designers work.At the same time, the scientific
literature does not yet offer much evidence to support the concerns of some ecolo-
gists that habitat corridors are detrimental in certain situations—for example, by
enticing animals into habitats where mortality risk from predators or road cross-
ings is higher, or by enhancing the spread of pests, wildlife diseases, and exotic
invasive species. Nevertheless, it is worth keeping these cautions in mind. En-
suring that any natural corridors consist of high-quality habitat with native vege-
tation would help minimize several of these concerns. A greater practical “cost”
The Ecology of Landscapes 103
of corridors is that limited resources will be spent to create corridors of marginal
conservation value rather than being used for more worthy projects.
20
We can gain additional insight on the value and optimal design of corridors
by once again thinking of the landscape from the perspective of different organ-
isms. One question is whether corridors are broadly effective—helping a wide
range of species—or whether they should be employed specifically to help a given
species of concern. In 1999, conservation biologist Andy Dobson and fourteen
coauthors representing a diversity of opinions answered this question by sug-
gesting that “the first step in the analysis of corridor capability [should be] the
selection of target species The idea of a generic landscape corridor—connec-
tivity for the sake of connectivity—is more aesthetic than scientific and will gen-
erally be dismissed in the hard light of scientific review.”
21
As Dobson and his

colleagues point out, corridors can be especially useful in carefully targeted con-
servation efforts, such as helping to sustain populations of species that are mi-
gratory or nomadic, or populations that are not likely to be viable in the long
term in established nature reserves.
Given the accumulating evidence that corridors can improve the viability of
populations, and given the great difficulty and expense of creating corridors after
a region becomes developed, it is wise to set aside corridors prior to or during the
development of a region. If we wait until we have comprehensive scientific data
about what kinds of corridors help what species, it may be impossible or at least
prohibitively expensive to “retrofit” a landscape with habitat corridors later. Ac-
cordingly, when a major project such as a road or shopping center is proposed
that would threaten habitat connectivity, planners and designers should presume
that the loss of connectivity would hurt the local biota and take steps to reduce
or mitigate this loss unless site-specific studies demonstrate otherwise. On the
other hand, when faced with the question of whether to spend limited conserva-
tion resources to protect a corridor at a specific location, planners and conser-
vationists would be wise to invest in ecological studies to determine whether the
proposed corridor would actually help the target species. If not, resources can be
redirected to address a more pressing need.
effects of human corridors
Numerous researchers have studied the effects of human corridors—particu-
larly roads—on populations and ecosystems. The most important ecological ef-
fect of human corridors is as a filter or barrier to the movement or dispersal of
native species. This and other effects of common human corridors are profiled
below and in Figure 6-5.
Roadkills occur in staggering numbers, with an estimated 1 million verte-
brates per day killed on roads in the United States alone.
22
The best way to reduce
104 THE SCIENCE OF ECOLOGY

this carnage is to limit the number of roads—an important goal for conservation-
minded planners and developers. Short of closing roads or not building them in
the first place, the most successful technique for mitigating roadkills is to install
fencing that restricts animal movement onto the road in conjunction with un-
derpasses or overpasses that allow animals to cross the road safely.
23
Underpasses
range from shallow tunnels for salamanders and other amphibians to wide
swaths of vegetation with the roadway elevated high above (see Figures 6-6 and
7-5e). Prefabricated underpasses (culverts) for amphibians and small mammals
are relatively inexpensive and could be incorporated into residential subdivisions
or commercial developments where a proposed road will divide a formerly con-
tiguous population or isolate feeding, breeding, or nesting habitats.
Overpasses can consist of raised arches over the highway (in some cases, up
to a few hundred feet wide) or bridges covered with natural vegetation that are
flush with the surrounding landscape and pass over a sunken roadbed (see Fig-
ure 7-5d). Any overpass or underpass system must be paired with effective fenc-
ing, berms, or other barriers to direct animals toward the crossing points. Wildlife
crossing systems should be designed around the needs of specific target species:
the largest animals of interest and the species most sensitive to the road barrier.
Accommodating the needs of these species should result in a system that works
for most other species. These needs should determine where the crossing struc-
ture is placed, whether it passes over or under the road, how large it is, and what
material is used for the surface.
In addition to reducing roadkills and enhancing wildlife movement, sensitive
road design should address the other major ecological impacts of roads, including
altered drainage and hydrology, pollutant runoff, and the spread of non-native
vegetation. Regarding vegetation, a recent effort to enhance roadside habitat has
involved planting native grasses, flowers, and shrubs rather than non-native
species.

24
This movement combines earlier objectives for roadside vegetation
management—stabilizing slopes, providing a “clear zone” for errant vehicles,
beautifying the roadside, and minimizing maintenance costs—with a new un-
derstanding of the potential ecological value of roadside habitats. For example,
Iowa’s Living Roadway Program encourages and offers grants for planting na-
tive species, including restored prairie communities, alongside the state’s roads.
Roadside managers and ecologists have found that the use of indigenous prairie
plants as well as less-intensive mowing and herbicide spraying regimens (or none
at all) actually reduces weed and erosion problems while improving habitat for
native grassland plants, birds, and insects.
25
State programs are not the only way to promote ecologically compatible
roadside management. Planners at the municipal or county level can encourage
or require the use of native roadside vegetation in new public and private devel-
The Ecology of Landscapes 105
Figure 6-5b. High-speed two-lane roads have the highest road-kill rates because
more animals attempt to cross these roads than try to cross superhighways. Many ani-
mals are lured to the road or roadside by the prospect of food, salt, a warm surface for
basking, or even the water that collects in puddles after a rainstorm. Road-kill rates are
expected to be high where a natural movement corridor intersects the road. Road mor-
tality is not likely to threaten populations of most rapidly reproducing animals but can
be a major factor for rare or less fecund species, especially large mammals. (Sources:
Patricia A. White and Michelle Ernst, Second Nature: Improving Transportation with-
out Putting Nature Second [Washington, DC: Defenders of Wildlife, 2003]; A. F. Ben-
nett, “Roads, Roadsides and Wildlife Conservation: A Review,” in Denis A. Saunders
and Richard J. Hobbs, eds., Nature Conservation 2: The Role of Corridors [Chipping
Norton, Australia: Surrey Beatty, 1991], pp. 99–117.)
Figure 6-5a. Of all corridor types, median-divided superhighways are the most likely
to inhibit animal crossings. Such barriers cause populations on either side of the high-

way to be isolated from one another, making each subpopulation more vulnerable to
extinction. The isolation effect applies to birds as well as insects, reptiles, amphibians,
and mammals. The edge effect of multilane highways extends anywhere from a few
hundred feet for many mammals and pollution-sensitive plants to a mile or more for
noise-sensitive grassland birds and other species. Most inhabitants of road edges and
medians tend to be edge species and exotics. (Sources: H J. Mader, “Animal Habitat
Isolation by Roads and Agricultural Fields,” Biological Conservation 29 (1984): 81–96;
Richard T. T. Forman et al., Road Ecology: Science and Solutions [Washington, DC:
Island Press, 2003].)
Figure 6-5c. A major effect of secondary roads is “taking up space” on the landscape.
Public roads and adjacent roadsides in the United States occupy roughly 27 million
acres, or 11 million hectares (1.2 percent of the U.S. land area), and the “road effect
zone” of degraded habitat near these roads encompasses almost one-fifth of the U.S.
land area. The use of open roadsides as a conduit for animal movement is the exception
rather than the rule, although road corridors do facilitate the spread of certain invasive
species. Even a narrow paved road can function as a barrier to movement for many in-
sect and small mammal species. Roads that separate amphibian breeding habitat from
adult habitat may have serious impacts on amphibian populations. (Sources: Forman et
al., Road Ecology; Richard T. T. Forman, “Estimate of the Area Affected Ecologically by
the Road System in the United States,” Conservation Biology 14, no. 1 [2000]: 31–35;
B. A. Wilcox and D. D. Murphy, “Migration and Control of Purple Loosestrife
[Lythrium salicaria L.] along Highway Corridors,” Environmental Management 13
[1989]: 365–70; Richard T. T. Forman and Lauren E. Alexander, “Roads and Their Major
Ecological Effects,” Annual Review of Ecology and Systematics 29 [1998]: 207–31.)
Figure 6-5d. While they are less of a barrier than paved roads for many species, nar-
row, unpaved roads still inhibit movement by many insects and small mammals.
Predators are known to travel along unpaved roads with little traffic. Even lightly used
forest roads promote human incursions into natural areas for hunting and logging and
help spread invasive species, whose seeds often hitch a ride on vehicles. Large mam-
mals, such as bear and elk, are very sensitive to road density. For this reason, some land

managers have proposed road closings in natural and seminatural areas to stabilize
populations of rare interior species. (Source: Bennett, “Roads, Roadsides and Wildlife
Conservation.”)
Figure 6-5f. Urban and suburban greenways combine multiple functions—habitat
protection, recreation, nonmotorized transportation, and opportunities for historic or
cultural appreciation—into a single corridor. Habitat is generally suitable mainly for
edge species due to the narrow width and intensive human use of the corridor. Most
greenways in developed areas have narrow spots or are intersected by roads, which
greatly limits their value for long-range wildlife movement. Riparian greenways can
help filter pollutants and excess nutrients, reduce erosion, and improve stream habitat.
(Source: Reed F. Noss, “Wildlife Corridors,” in Daniel S. Smith and Paul C. Hellmund,
eds., Ecology of Greenways [Minneapolis: University of Minnesota Press, 1993].)
Figure 6-5e. Rail corridors are rarely completely devoid of native species, but the
habitat value of these areas varies greatly depending on how they are managed. Rem-
nant strips of natural vegetation are the most favorable for native species, and rail cor-
ridors are often more likely than roads to exhibit such “benign neglect.” For example,
in agricultural areas of the Midwest, rail corridors contain some of the last remnants of
native prairie and have therefore been a critical source of seeds for native plants used
in prairie restoration projects. Active and abandoned rail corridors in urban areas can
be important ecologically because they are some of the few unmanaged areas within a
heavily managed matrix.
Figure 6-5g. Unlike roads, trails and paths are often used by mammals as conduits
for movement. However, heavy use by humans or even limited use by dogs (which
leave scent marks) sharply reduces use by wild animals. Well-defined narrow trails
have less impact than wide or braided trails because human activities are less dispersed
and animals can learn to avoid them. Therefore, in sensitive nature reserves, land man-
agers may want to confine most human use (and all use by dogs) to a portion of the
site near the edge. (Source: Richard T. T. Forman, Land Mosaics: The Ecology of Land-
scapes and Regions [Cambridge: Cambridge University Press, 1995], p. 174.)
Figure 6-5h. As with roads, utility corridors (power lines, gas and oil pipelines, and so

forth) contain mainly edge species. Most utility corridors are kept open by regular dis-
turbance from humans, such as cutting or herbicide spraying. Studies show that they
inhibit crossing by many mammal, bird, amphibian, and insect species. Ecologically
sound management might involve planting with native herbaceous and shrub species
that would require less frequent maintenance, provide better habitat for native ani-
mals, and create less of a barrier to movement. Also, curvilinear, “soft” edges might
encourage animal movement into and across the corridor. (Sources: Forman, Land
Mosaics, p. 174; H. H. Obrecht III, W. J. Fleming, and J. H. Parsons, “Management of
Powerline Rights-of-way for Botanical and Wildlife Value in Metropolitan Areas,” in
Lowell W. Adams and Daniel L. Leedy, eds., Wildlife Conservation in Metropolitan
Environments [Columbia, MD: National Institute for Urban Wildlife, 1991], p. 255.)
opments, while engineers and landscape architects can propose the use of native
grass or shrub ecosystems as aesthetically pleasing and low-maintenance alter-
natives to monocultures of non-native grasses.
Land Mosaics, Land Transformation, and Implications
for Planning
Taken as a snapshot at a single point in time, the land displays a mosaic, or quilt-
like, pattern of patches, corridors, and matrix. This mosaic is created by variability
in the environment (e.g., soils, moisture, and topography), natural disturbances,
and human activity. However, viewed 10, 50, or 100 years later, the mosaic is
likely to look different. Two processes are responsible for this change.
In the absence of human activity, the natural processes of disturbance and
succession discussed in Chapter 4 result in shifting mosaics, in which individual
patches change from early successional to late successional vegetation and vice
versa but the landscape as a whole remains in general equilibrium (see Figure
6-7). Since different species rely on different successional stages for light, nu-
trients, food, and shelter, it is important that there be at least some patches at
110 THE SCIENCE OF ECOLOGY
Figure 6-6. Salamanders use this tunnel to cross under a road during their annual
migration to breeding ponds. Note the fencing and concrete “funnel” in the fore-

ground of the photo, which guide salamanders toward the underpass and prevent them
from accessing the road surface.
every successional stage at any given time. For example, in the forests of north-
ern New England and eastern Canada, moose (Alces alces) find most of their food
in young hardwood stands where the forest was recently cut or damaged by wind
or ice, whereas the moss Neckera pennata occurs only in late successional forests
in this landscape.
26
A landscape lacking either of these forest types could not sup-
port all native species.
The Ecology of Landscapes 111
Figure 6-7. Even in the absence of human
intervention, landscapes change as a result of
succession and disturbance. This diagram
shows the same forested landscape over time,
with fifty years passing from one panel to the
next. Individual forest patches mature from
young (white) to middle-aged (gray) to old
(black), while natural disturbances create new
young patches from older ones. Over time,
however, the landscape as a whole remains a
mosaic with forests of all different ages
represented.
Young Forest
Middle-Aged Forest
Older Forest
Compared to a natural mosaic where succession and disturbance follow each
other in a continual cycle, a human-influenced mosaic is more likely to change
directionally as the matrix of natural habitat is interspersed with more and more
human land uses. Understanding how this happens is useful for planners and de-

signers. Land transformation can occur through several different processes: per-
foration, dissection, fragmentation, shrinkage, and attrition (see Figure 6-8).
27
All
five of these processes are often lumped under the term fragmentation in com-
mon parlance, but there are important differences from an ecological standpoint.
Perforation occurs when scattered houses are built within natural habitat or
when remote patches of forest are clear-cut, for example.This process rapidly in-
creases the amount of edge and decreases the maximum size of an uninterrupted
interior habitat patch. For example, if just ten houses are scattered throughout a
remote, forested, 1,000-acre (400 ha) section of a town, roughly 110 acres (45 ha)
of edge habitat will be created, assuming a 300-foot (90 m) edge width.
28
Perfo-
ration is most troublesome for species that require large patches of interior habi-
tat. At least initially, perforation is not likely to subdivide natural populations,
because the matrix of natural habitat remains continuous.
Dissection is caused by roads and other human corridors carving continuous
swaths through the matrix. As discussed above, different types of human corri-
dors create barriers for some species but not for others. If individuals of a species
are unable to cross the corridor, then the population will be subdivided. This may
occur even though less than 5 percent of the landscape has been directly altered.
Dissection also creates a large amount of edge relative to the amount of land that
is directly altered.
Fragmentation and shrinkage occur when patches of natural habitat become
discontinuous from one another. When this happens, the matrix may change
from being natural habitat to a human land use. At this point, many if not most
interior species will probably have been lost, and many other species will have
been divided into metapopulations or reduced to unsustainably small populations
that may soon disappear. Nevertheless, the landscape may still provide habitat

for generalist species, species with small home-range requirements, migratory
birds, and other species that can use stepping stones of natural habitat in mov-
ing from one core habitat area to another.
Attrition is the final stage of land transformation, during which the residual
patches of natural habitat are lost completely. At this stage, even the amount of
edge habitat diminishes rapidly, and the biota are likely to be limited to those
species that can tolerate human land uses and activities.
The study of land transformation reveals two useful principles for planners
and developers. First, the greatest impact to sensitive native species usually oc-
curs early in the land transformation process—by the time one-quarter of a com-
112 THE SCIENCE OF ECOLOGY
Figure 6-8. This series of diagrams illustrates the various land transformation
processes that occur as a result of human settlement. In sequence, they show an unin-
habited forested landscape (a), dissection (b), perforation (c), fragmentation (d), shrink-
age (e), and attrition (f).
A B
C D
FE
munity’s land has been developed.Thus, ecologically based planning should begin
immediately, rather than waiting until substantial growth has already come to
a town, county, or region. Second, dispersed development is almost always more
detrimental to natural communities than is a comparable amount of concentrated
development because it accelerates all five land transformation processes: perfo-
ration, dissection, fragmentation, shrinkage, and attrition. This principle provides
ecological support for clustering development at the site level and especially at
the municipal and regional level so that large remnant patches of native vegeta-
tion can be retained.
29
Having reviewed the major concepts of landscape ecology, we can now ask
two questions of great relevance to planners and developers: where and in what

sequence should land be developed to maximize ecological values? Landscape
ecologists Richard Forman and Sharon Collinge have proposed conceptual an-
swers to these questions, and the following discussion is based on their “spatial
solution” to land use planning.
Where is the best place, ecologically, to situate any given land use, such as
a new housing development, road, shopping center, farm, or nature reserve? Al-
though the answer to this question depends on place-specific variables, landscape
114 THE SCIENCE OF ECOLOGY
Box 6-1
”Indispensable Patterns” for Biological Conservation
1. Large natural patches. Large patches are the only way to protect interior species and species
with large home ranges. Large patches are also more likely to allow for a shifting mosaic in
which natural disturbances do not affect all the land at once and, thus, several successional
stages (with their associated biotic communities) are represented at any given time.
2. Vegetated riparian corridors. Naturally vegetated streamsides are essential for protecting
many aquatic species important to conservation.
3. Connectivity between large patches. The landscape must provide functional connectivity for
species of conservation interest—that is, linkages that these species can use for home range
movement, migration, and dispersal. Wide, continuous corridors are most likely to serve this
function, but stepping stones in a moderately suitable matrix will suffice for many species.
4. Natural remnants in human-dominated areas. Within urban or agricultural landscapes, three
types of natural remnants should be protected, in order of descending priority: (1) areas of
especially high conservation value, such as microhabitats that are rare throughout the land-
scape; (2) landscape types that provide essential ecosystem services, such as flood control;
and (3) remnants of the former natural matrix that provide edge species habitat and human
access to nature.
Source: Based on discussion in Richard T. Forman, Land Mosaics: The Ecology of Landscapes and Regions (Cambridge Uni-
versity Press, 1995), p. 452.
The Ecology of Landscapes 115
Figure 6-9. The aggregate-with-outliers model, illustrated here, has been proposed as

one way to incorporate biological conservation and human land uses at the landscape
scale (tens of miles or kilometers across). (Based on Richard T. T. Forman and Sharon
K. Collinge, “The ‘Spatial Solution’ to Conserving Biodiversity in Landscapes and Re-
gions,” in R. M. DeGraaf and R. I. Miller, eds., Conservation of Faunal Diversity in
Forested Landscapes [London: Chapman and Hall, 1996], pp. 537–68.)
ecology can offer a useful generic answer, which can then be adapted to the plan-
ning or design questions at hand. At the scale of landscapes, four “indispensable
patterns” of natural vegetation must be maintained in order to protect native
species and natural processes.
30
These patterns are discussed in Box 6-1.
The aggregate-with-outliers model is one possible way of incorporating
these four patterns into land use plans.
31
This design proposes that major land
uses—such as natural vegetation, agriculture, and urban development—should
generally be aggregated to maximize large-patch benefits. However, small out-
lying patches should also be created to provide for edge habitat, reduce the risk
of catastrophic disturbances or pest outbreaks affecting a single patch, increase
genetic variation, and provide opportunities for human appreciation of nature.
Connectivity for native species should be provided by natural corridors as well
as small patches, which can function as stepping stones (see Figure 6-9).
The aggregate-with-outliers model does not apply in exactly the same ways
to both natural vegetation and urban areas. As we discuss throughout this book,
people in urban and suburban areas benefit from having small patches of natu-
ral vegetation sprinkled throughout the developed landscape. In contrast, small
patches of urban areas do not improve the functioning of native ecosystems,
which are best left in large patches. Once again, these principles (and Figure 6-9)
offer a generic solution for ecologically based land use planning—a solution that
must be refined based on the details of each place.

What is the ecologically optimal sequence of land transformation? Planners
use many techniques to influence the sequence of land transformation—that is,
the order in which land is developed or altered. Zoning maps, infrastructure
investment programs, urban growth boundaries, development phasing ordi-
nances, incentives, and subsidies all affect the sequence of land transformation.
An ecologically based approach to land transformation would maximize each of
the four indispensable patterns on the landscape for as long as possible. Thus, de-
velopment should be aggregated on less ecologically important lands, reserving
natural areas that are large and relatively round for as long as possible. Within
the land that is converted for agriculture or development, remnant small patches
and corridors should also be set aside. Figure 6-10 illustrates how this more eco-
logically favorable land transformation process might look at 20 percent, 50 per-
cent, and 80 percent of total buildout. Even at the 80 percent stage, the landscape
may still support many species of conservation interest in the remaining large
patch and perhaps in some of the smaller remnant patches. This is a radically dif-
ferent outcome from conventional urban development (or agriculture), in which
nature is relegated to leftover scraps or small parks selected with little regard for
regional ecology.
How practical or implementable are these “optimal” land use scenarios of-
fered by the field of landscape ecology? At first glance, the idea of identifying
lands for development, agriculture, and natural habitat at the scale of towns or
counties appears to be at odds with property rights concerns, not to mention most
current practice. However, directing land use in this manner can be accomplished
using zoning and planning tools, as has been done successfully in some notable
instances. For example, urban growth boundaries (e.g., Portland, Oregon) and re-
gionwide transfer of development rights programs (e.g., the Pinelands in New
Jersey) are both essentially techniques for achieving an aggregate-with-outliers
land use pattern at the landscape scale (see Chapter 10 for further discussion of
these tools).
Conversely, large-lot zoning (roughly 1 to 40 acres, or 0.4 to 16 hectares, per

lot) runs counter to these ecologically based land use solutions because it cre-
ates a fine-scale intermingling of developed, agricultural, and natural lands that
eliminates the large-patch benefits to be gained by aggregation. While some eco-
logical values can be retained on large residential lots (see Box 10-1 for an enu-
meration of these), even 35-acre (14 ha) “ranchette” house lots in the western
United States were found to have significantly fewer native birds and predators
and significantly more introduced predators and plants than nearby ranch lands
and nature reserves.
32
Because it is so spread out, ranchette development degrades
far more natural habitat than would an equal number of houses in an ordinary
suburban development.
116 THE SCIENCE OF ECOLOGY
Even at the scale of individual sites, the ecological benefits of aggregating or
clustering development as opposed to distributing it evenly across the site are
apparent. For example, if a 640-acre (1 square mile, or 260 ha) tract of land is di-
vided into sixteen 40-acre house lots—a common pattern in the West—76 per-
cent of the tract will be affected by development, assuming a 650-foot (200 m)
disturbance radius (edge effect) around the houses. However, if the houses are
The Ecology of Landscapes 117
Figure 6-10. This time series of three
sketches showing a community at 20 percent
(a), 50 percent (b), and 80 percent (c) of
maximum buildout illustrates how the land
transformation sequence can be improved
to maximize conservation values. Whereas
the typical sequence of community develop-
ment results in early and extensive habitat
degradation (as shown in Figure 6-8), this
improved sequence retains the maximum

amount of large habitat patches and corri-
dors at each stage of development. (Based on
Richard T. T. Forman and Sharon K. Collinge,
“The ‘Spatial Solution’ to Conserving Bio-
diversity in Landscapes and Regions,” in
R. M. DeGraaf and R. I. Miller, eds., Conser-
vation of Faunal Diversity in Forested
Landscapes [London: Chapman and Hall,
1996], pp. 537–68.)
A
B
C